Image sensors are semiconductor devices used to convert an optical image into an electrical signal, and are typically classified into a charge coupled device (CCD) image sensor and a CMOS image sensor (CIS).
According to the conventional technology, a photodiode (PD) is formed by implanting ions into a substrate. However, as the size of the PD is gradually decreased such that the number of pixels is increased without the increase of a chip size, an image quality may be degraded due to the area reduction of a light receiving part.
In addition, since a stack height has not decreased in a corresponding manner to the area reduction of the light receiving part, the number of photons input into the light receiving part may be decreased due to the diffraction of light called “Airy Disk”.
As an alternative method to overcome the above problems, the PD may be deposited by using amorphous silicon (Si). In addition, after a readout circuitry is formed on a Si-substrate through a wafer-to-wafer bonding scheme, the PD may be formed on the readout circuitry (hereinafter, referred to as “three-dimensional image sensor”). In this case, the PD is connected with the readout circuitry through a metal line.
Meanwhile, according to the conventional technology, after two wafers are bonded to each other, signal lines must be formed on the wafer formed with the PD. To this end, after a contact is formed in a logic wafer, the logic wafer is bonded to the wafer formed with the PD. Then, silicon (Si) of the wafer formed with the PD is removed except for the PD area. In this case, high contact resistance may exert a bad influence upon the sensitivity of the image sensor. In other words, in the conventional structure, the sensitivity reduction of the image sensor is caused by high contact resistance.
According to the conventional technology, since a source and a drain of a transfer transistor are heavily doped with N-type dopants, charge sharing may occur. If the charge sharing occurs, the sensitivity for an output image is degraded, so that image errors may be caused.
According to the conventional technology, since photo-charges do not smoothly move between the photodiode and the readout circuit, dark current may occur, or saturation and sensitivity may be degraded.